The proposal's broad, long-term objective is to improve the health of millions of Americans with or at risk for non-alcoholic fatty liver disease (NAFLD). The immediate goal is to technically validate magnetic resonance imaging (MRI) determined proton density fat fraction (PDFF) as a biomarker of liver fat by demonstrating: Aim 1. Accuracy against an independent, definitive reference standard. We will determine and compare the accuracy of PDFF as measured by two MRI techniques, using tissue triglyceride concentration as the reference standard. We hypothesize that PDFF will accurately predict TG concentration. We will identify the more accurate technique. Aim 2. Precision with high repeatability and reproducibility. We will assess the repeatability within subjects and reproducibility across sites, scanners and readers of PDFF values as measured by two MRI techniques. We hypothesize that PDFF will be repeatable and reproducible. We will identify the more repeatable and reproducible technique. Aim 3. Clinical and research value. We will use PDFF as a biomarker to define the relationship between weight loss and liver fat in obese adults undergoing gastric banding, demonstrating its utility as an objective tool for repeated measures of liver fat content. We hypothesize that PDFF will decrease monotonically in response to weight loss. Factors that modulate this relationship will be identified. Significance: NAFLD is the most common chronic liver disease of adults and children in the United States. It may progress to cirrhosis and contribute to the development of cardiovascular disease and type 2 diabetes. Liver biopsy, the current clinical gold standard to assess liver fat and diagnose NAFLD, is invasive. This limitation has hindered progress in the diagnosis, treatment, and prevention of NAFLD. A valid non-invasive biomarker of liver fat content is needed for clinical care and research. New breath-hold MRI methods refined at the University of California, San Diego and the University of Wisconsin address the sources of errors in conventional MRI for measuring liver fat. These new methods permit rapid estimation of PDFF, the fraction of mobile protons in liver attributable to fat. Preliminary data suggests that PDFF accurately predicts liver TG concentration, the biochemical measure of liver fat content; but rigorous validation is required to establish MRI-determined PDFF as a reliable and useful biomarker. Research design and methods: We plan a two-site prospective cohort study. The two-site design provides unbiased comparison of competing techniques, adequate patient enrollment, and the ability to compare performance characteristics across sites. We will enroll 140 obese adults (70 per site) participating in a surgical weight loss program (laparoscopic adjustable gastric banding). Enrolled subjects will have variable baseline liver fat content; have multiple intra-operative biopsies providing a definitive reference standard with which to compare PDFF measurements; and lose weight and liver fat at variable rates. PDFF will be measured using two MRI techniques at multiple time points before and after surgery.